Missing Is Not Zero
A null address, unknown contract end date and zero invoice total mean different things.
For Data & Technology Analysts
Data Analyst Tools for Reports People Can Trust
Connect approved data, check what changed and explain the result before sharing it. The right data analyst tools help you build useful reports and show where the numbers came from.
Mimasa AI brings prepared data, dashboards and reviewed analysis into one governed workspace. Start with one dataset. Check its quality and refresh date, then publish the view your audience is allowed to see.
Defined checks · Source dates · Traceable changes · Human review
01 · Approved Source
Order Feed · As Of 09:00 UTC
02 · Defined Check
Order ID Present · 960 / 1,000
03 · Dated Dashboard
Updated 10:00 UTC · Review Needed
Reviewed Note
Forty rows need review. The source may still be incomplete, even though the job finished.
Analyst Method
What Should a Data Analyst Workflow Check?
A data analyst workflow starts before the chart. Decide which source is authoritative, what each field means and when records should arrive.
Data Extraction & Connectivity handles approved source access. No-Code Data Transformation supports preparation where configured. The analyst checks the result.
- 01
Source
Which approved systems or files supply the records?
- 02
Shape
Which keys, dates and types should be present?
- 03
Rules
Which records should pass, fail or wait for review?
- 04
Refresh
When did the source and report last update?
- 05
Impact
Which dashboard or report uses the changed field?
- 06
Finding
What is known, uncertain and ready for review?
Defined Quality
Forty rows need review. This does not prove an upstream failure or missing customer order. It is not a customer outcome.
Build a Data Quality Dashboard With Clear Rules
A data quality dashboard should show useful checks, not one unexplained health score. Start with completeness, validity and freshness. Assign every exception an owner.
For a data quality metrics dashboard, place the definition beside the result. A data quality monitoring dashboard compares stable measures over time.
| Check | Plain Meaning | Analyst Must Define |
|---|---|---|
| Completeness | Required values are present | Required fields and eligible rows |
| Validity | Values meet an agreed rule | Accepted format, range and exceptions |
| Uniqueness | Keys expected to be unique are unique | Key and legitimate-repeat treatment |
| Freshness | Expected records arrived on time | Schedule, arrival time and tolerance |
| Consistency | Two approved views agree | Scope, join key and period |
| Measure | Fictional Result | Defined Basis |
|---|---|---|
| Rows Received | 1,000 | Feed received for 26 September 2026 |
| Nonblank Order ID | 960 | 960 ÷ 1,000 = 96% completeness |
| Missing Order ID | 40 | Kept for review; not silently dropped |
| Source As Of | 09:00 UTC | Arrival time, not proof all orders arrived |
BI Validation
Validate Inputs Before a BI Report Goes Live
BI analyst tools compare a draft view with its source and previous report. Business intelligence analyst tools should expose filters, join keys and definitions.
A data validation dashboard shows which checks passed, failed, lack data or are not configured. A data analyst dashboard discloses those states when they affect a number.
Good dashboard data quality comes from checking the figures people use. See Visualisation & Dashboards and Data Governance.
Before Approval
- Did a field change type or meaning?
- Did an unexpected join filter out records?
- Does a late correction alter the previous period?
- Do checks and charts use the same dataset version?
Dataset Profile
Profile a Dataset Without Hiding Its Gaps
Profiling is a first look at row count, field types, missing values and unusual ranges. A data profiling report makes those checks readable.
When supported, a data profiling dashboard can compare a new file with the previous one. It must label each difference and likely impact.
Retain Exceptions
Keep rejected, late, duplicate and unmatched records available to authorised reviewers.
Describe Schema Change
A changed postal-code field may affect a region chart but not a product total.
Compare Dated Versions
Intelligent Data Snapshots support dated comparisons where configured.
Configured Signals
Monitor Freshness and Pipeline Errors in Context
Data freshness monitoring asks whether an approved source arrived as expected. Source freshness, job completion, dataset completeness and dashboard update time are separate facts.
A data pipeline monitoring dashboard can compare configured jobs, arrivals and errors. An ETL monitoring dashboard adds extract, transform and load stages when history is available. ETL data quality checks compare expected input and output counts.
| Signal | What It May Mean | Next Check |
|---|---|---|
| Late Source Arrival | New records may be delayed | Check schedule and last successful arrival |
| Fewer Rows | Feed, filter or business volume changed | Compare source count and accepted exclusions |
| Changed Field Type | An upstream format may have changed | Check schema and affected transforms |
| Job Failure | A configured step did not complete | Inspect error, owner and last valid output |
For data pipeline error monitoring, show only failures emitted by configured jobs. Data observability metrics may include freshness, volume and schema change. A data observability dashboard is a configured source-health view, not a promise to replace a specialist platform.
Known Path
Trace a Changed Number Back to Its Source
Data lineage analysis follows one number through a report, prepared dataset, transformation and approved source.
If a region total falls, check the filter, mapped field, source file and refresh date. Record every path you cannot verify. A plausible explanation is not a proven cause.
- 01
Observed
Name the report, measure and period that changed.
- 02
Checked
List the source records, filters and transforms reviewed.
- 03
Unknown
State which feed or mapping still lacks evidence.
- 04
Next
Ask the owner to confirm before redistribution.
Reviewed Finding
Turn a Check Into an Analyst Finding
Data analyst reporting tools should keep the method beside the finding. State what changed, which source supports it and who reviews it.
Use Insights, Analysis & Reporting for a draft based on approved figures. Review the wording before distribution. For repeated decisions, see Data & Decision Intelligence.
Illustrative Finding — Fictional Data:Forty of 1,000 rows in the 26 September feed lack an order ID. Completeness is 96% for this feed.
We have not checked whether those rows are expected corrections. The dataset owner will review them before the region report is shared.
Authorised Audience
Share the Dashboard With the Right Audience
Show an as-of date on each view. If refreshed data changes the number, flag earlier commentary for review.
Private
Draft checks, failed rows and working notes
Review: Source access and purpose
Internal Team
Reviewed trends and assigned issues
Review: Permissions and exposed record details
Public, Only If Enabled
Cleared aggregate methods and findings
Review: Owner approval, privacy and current feature support
Version controls, public publishing, author profiles and reputation features require verified app support and data-owner permission.
Connected Questions
Explore Related Analyst and Business Questions
This role checks whether reporting inputs are fit for use. Specialist pages own the decision that follows.
Financial Reconciliation Automation owns controlled financial matching. Banking Regulatory Reporting Automation owns bank evidence and sign-off.
FAQ
Data & Technology Analyst FAQs
Clear answers about quality checks, freshness, lineage, monitoring and reviewed dashboard sharing.
Get Started
Make the Next Report Easier to Trust
Choose one recurring dataset and one dashboard. Define a few checks, inspect the changes and share a reviewed finding.